The Utah Medical Licensing Board recently voiced concerns with a pilot program that would allow Utahns to see an AI-generated “doctor.” This “doctor” could renew prescriptions for “low-risk” and routine medications, saving consumers money and time (no more trips to the doctor just to get a refill on a prescription you’ve had for a year).
The Board, however, felt that unsupervised AI creates unwarranted risks to public health and raised those concerns in a letter to the state’s Department of Commerce, urging the immediate suspension of the program pending its review by the Board. When the Department declined to take this step, the letter became public and the press reported on it.
I have served as a member of the Board since October 2024. I should clarify that unlike most of the Board, I am not a medical doctor. I serve as a “public member,” since my day job is associate professor of American history at Brigham Young University.
Eleven members of the Board signed the letter objecting to the AI pilot program. I was one of three who did not.
To be clear, many of the concerns raised by my fellow members struck me as thoughtful and consistent with the Board’s mission. The Board rightly argued that the state should be cautious in adopting AI, since experience tells us that new technology nearly always brings unforeseen risks. Moreover, the state erred in leaving the Board out of the development process until after the program went into its pilot phase. It makes little sense to convene a body of experts only to not consult with them in a timely way.
At the same time, the Board overreacted to the state’s pilot program. Yes, AI will almost certainly create unforeseen problems as it is adopted into medical practice. But it will not go away.
In my view, it is better to do what the state has done in creating a supervised pilot program that can reveal those problems early in the adoption process. It’s also better for the Board to watch over the adoption process rather than kill the pilot before it has a chance to produce any evidence.
In addition, in its letter to the Commerce Department, the Board implied that the pilot program privileged financial concerns over public health. This struck me as missing the point.
As a student of economic and political history, I know that we cannot so simply divide the cost of care from the quality of care. Making medical services more efficient means making them better and more available.
The Board rightly worries about the risks involved in adopting a new technology, but it should also acknowledge that AI has the potential to provide more quality care for a greater number of Utahns should it deliver on its potential. Indeed, the benefit of a pilot is precisely to develop more data with which to weigh the potential risks and rewards of the new technology.
Over the last decades, the cost of medical care has grown significantly faster than our overall economy. In 1960, our nation spent 5% of our GDP on health care; today, we spend about 18%. This shift in spending results in part from the fact that the health care industry has lagged behind the rest of the economy in productivity gains.
According to a 2016 study from the Centers for Medicare & Medicaid Services, between 1990 and 2013, the productivity increases at American hospitals were half those of the rest of the economy. Comparatively speaking, we pay more for our health care to receive less.
There are many complicated historical factors that explain this inefficiency, ranging from the rising costs of malpractice insurance to the extraordinary administrative costs that follow our complicated and overlapping system of public and private insurance.
Most histories of health care do not explore all the reasons for the costs of the system and instead tell the story of the many dramatic political fights involved in creating a government-run national health care system. Franklin D. Roosevelt first contemplated this as part of creating Social Security. Subsequently, Harry Truman, Lyndon B. Johnson, Bill Clinton and Barack Obama all hoped to move the country closer to the socialized, single-payer system common in Western Europe.
In each case, the medical profession managed to thwart or at least reshape these presidents’ intended outcomes, usually with the help of conservatives and libertarians. In opposition to nationalized health care, this coalition argued that doctors and markets should decide who gets what treatment and when. So far, the conservative coalition has largely, but not completely, prevailed.
The current controversy departs from that nearly century-long configuration of issues and alliances. AI does not threaten to centralize and bureaucratize medicine. Rather, AI threatens to marketize it by creating a new sector that would compete with doctors in providing medical services.
If AI reaches its potential, multiple virtual systems will compete not only with doctors but also with each other. This should lower prices while generating greater efficiencies and, in the utilitarian sense, provide the best care for the greatest number.
For the medical profession, though, this could create a challenging situation. As AI squeezes its profits, the profession might find itself fighting a two-front political battle: On one side, the political left will continue to advocate for a European-style single-payer system; on the right, free-market advocates will press for technology to assume ever more tasks traditionally reserved for doctors.
In other words, AI threatens to split the political coalition that served the medical profession so well for so long. This short letter from the Medical Board, narrowly focused as it is on a single pilot program, might turn out to be an early skirmish in a long and drawn-out fight that will decide the fate of the medical profession. Utah might be the first front in that war.